I wonder if it would be practical to do the kind of layout randomisation that Emery Berger discusses here. They found that the -O3 optimisation level was not statistically better than -O2 for C++ (with clang, I think)
Would be interesting to see how compilers for other languages fare.
I wonder if it would be practical to do the kind of layout randomisation that Emery Berger discusses here. They found that the -O3 optimisation level was not statistically better than -O2 for C++ (with clang, I think)
Would be interesting to see how compilers for other languages fare.